From: "Paul Keinanen" <keinanen@sci.fi>
To: rsgb_lf_group@blacksheep.org
Subject: LF: Re: Signalling margins and Shannon, shud we move on?
Date: Tue, 20 Feb 2001 05:53:59 +0200
Message-ID: <fro39t4thgfu1vq28at2p6mpusbupoicod@4ax.com>
References: <65AECDF1F89AD411900400508BFC869F0D75CC@pdw-mail-1.dera.gov.uk> <3A9160B5.D9C9DE4D@usa.net> <002801c09acb$30ae6cf0$0a00a8c0@ThreeLakes.ca>
In-reply-to: <002801c09acb$30ae6cf0$0a00a8c0@ThreeLakes.ca>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 8bit
Precedence: bulk
Reply-To: rsgb_lf_group@blacksheep.org
Sender: <majordom@post.thorcom.com>

On Mon, 19 Feb 2001 18:01:13 -0500, "Larry Kayser"
<kayser@sympatico.ca> wrote:


>This needs to be carefully thought through - I for one am reluctant to bring
>up issues about 300 second data bits simply because I know that many of the
>participants on this reflector are not even happy about 30 second visual CW
>bits.  I also am not keen to get into discussions that I know full well are
>making other users of the facility unhappy.

Extending the bit time might be a good idea if the propagation
conditions would remain constant as in space communication. However,
in variable propagation conditions, if the message takes hours to
transfer, some bits would get through quite strongly, while others
would be wiped out completely. Interleaving and ECC would help against
this, but again this would extend the transmission period further.

How about some Memory-ARQ style system ? 

Instead of standardising the bit rate, the message would be repeated
at an exactly defined ratio (eg. once every 600 s) and the signal from
various frames would be summed at  the receiving end, integrating the
power for a specific bit from multiple frames.  The end of the frame
(say 60 s) could be a acknowledge period, so when the receiver think
that the message has been decoded (e.g. by visual decoding), the
recipient would start to send ACK (possibly with a lower data rate)
during each acknowledge period. At the original transmitting site this
ACK message would finally accumulate to an ACK code and the original
transmitter could then stop transmitting, cutting the transmission
time to a minimum.

With a symmetrical link (same ERP and same background noise levels)
the recipient could event start sending  ACK as soon as the message
framing has been established (while still many data bits are unclear),
since during the time the ACK message has accumulated sufficiently,
most likely the original message would be clear enough :-).

Anyway, stations close to the transmitter site could decode the
message within a few minutes and thus now what is going on, without
having to wait for hours to se what that mysterious transmission is
all about.

Paul OH3LWR